@archanaahlawat Profile picture

Archana Ahlawat

@archanaahlawat

product, AI guardrails @ Dynamo AI || prev: research & eng @PrincetonCITP, @Microsoft, @JustFuturesLaw, @DukeU.

Similar User
⚡ reboot ⚡ photo

@reboot_hq

CharacterHub photo

@CharacterHub

Divya Siddarth photo

@divyasiddarth

Matthew Sun photo

@MatthewDSun

spencer chang photo

@spencerc99

jasmine sun photo

@jasminewsun

Saffron Huang photo

@saffronhuang

will photo

@will__ye

Max Langenkamp photo

@mslkmp

Gretchen Krueger photo

@GretchenMarina

Markus Anderljung photo

@Manderljung

shira photo

@AbramovichShira

Benjamin Laufer photo

@BenDLaufer

lucas gelfond photo

@gucaslelfond

jessica dai photo

@jessicadai_

Archana Ahlawat Reposted

personal news: I quit my job! I'm going to write full-time!

Tweet Image 1

Archana Ahlawat Reposted

Most real-world AI applications involve human-model interaction, yet most current safety evaluations do not. In a new paper with @saffronhuang @_lamaahmad @Manderljung, we argue that we need evaluations which assess human-model interactions for more accurate safety assessments 🧵

Tweet Image 1

Archana Ahlawat Reposted

👋Hello World! We’re Digital Witness Lab. We build privacy-respecting tools to access hard-to-reach data and empower public interest investigations.


Archana Ahlawat Reposted

Independent evaluation of the risks of foundation models is a vital means for transparency and instrument for accountability. While possible today, uncertainties and the lack of protections reduces its efficacy. New work led by @ShayneRedford to address sites.mit.edu/ai-safe-harbor/

Tweet Image 1

Archana Ahlawat Reposted

🚨New paper🚨 Black-Box Access is Insufficient for Rigorous AI Audits AI audits are increasingly seen as key for governing powerful AI systems. But to be effective, audits need to be high-quality, and to produce high-quality audits, auditors need access.🧵 arxiv.org/abs/2401.14446

Tweet Image 1

Archana Ahlawat Reposted

the thing about AI that people don't understand is that it's got all these risks. but also ! all these opportunities. not to mention the risks. but ! think of the opportunities. but the risks :( but the opportunit


Archana Ahlawat Reposted

If you are a public figure and tell your followers that “big new risks from advanced AI are fake”, you are wrong. Not only that, you’ll be seen to be wrong *publicly & soon*. This is not an “EA thing”, it is an oncoming train and it is going to hit you, either help out or shut up


Archana Ahlawat Reposted

A brilliant paper from my @PrincetonCITP colleagues analyzing the knowledge creation and sharing practices (forecasting, EA-funded prize competitions, forums) in the AI safety community that enabled this niche and contested set of ideas to go mainstream: drive.google.com/file/d/1HIwKMn…

The UK AI Safety Summit kicks off this week, highlighting the rapid rise of the field of AI safety and its influence on industry, academia and policy. In a new paper, Shazeda Ahmed, @archanaahlawat, @aawinecoff, @m0namon & I explain how this once-niche field became mainstream.🧵

Tweet Image 1


Archana Ahlawat Reposted

🧵Announcing GPQA, a graduate-level “Google-proof” Q&A benchmark designed for scalable oversight! w/ @_julianmichael_, @sleepinyourhat GPQA is a dataset of *really hard* questions that PhDs with full access to Google can’t answer. Paper: arxiv.org/abs/2311.12022

Tweet Image 1

Archana Ahlawat Reposted

Now is probably the time to announce that I've been writing a book about @OpenAI, the AI industry & its impacts. Here is a slice of my book reporting, combined with reporting from the inimitable @cwarzel Inside the year of chaos that led to this weekend. theatlantic.com/technology/arc…


Loading...

Something went wrong.


Something went wrong.